NEW: Kimi K2.7 Code HighSpeed is live - up to 260 tok/s, 6× faster. Access via API →
NOW AVAILABLE · JUNE 15, 2026 · HIGHSPEED MODE

Kimi K2.7 Code HighSpeed

Up to 260 tokens per second. Six times faster than standard. The same K2.7 Code model - Moonshot AI's most capable coding model to date - now served at extreme throughput for teams who need real-time agentic speed. 30% fewer reasoning tokens than K2.6. Now rolling out to Kimi Code Beta, API developers, and Business users.

⚡ Announced June 15, 2026 🧠 Same model as K2.7 Code 🔗 kimi-k2.7-code-highspeed 📜 Modified MIT License
// Peak throughput · Kimi K2.7 Code HighSpeed
0 tok/s
Peak · Short-Context Tasks
~180tok/s median coding
~260tok/s peak short-ctx
vs standard K2.7
260Peak tok/s
Faster vs std
256KContext
−30%Thinking tokens
MITLicense
01 - WHAT IS K2.7 CODE HIGHSPEED

Same Power. Extreme Speed.

Kimi K2.7 Code HighSpeed is a high-throughput serving variant of Kimi K2.7 Code - the identical model weights, optimized at the infrastructure layer for maximum output speed. There is no capability trade-off: HighSpeed uses the same trillion-parameter MoE architecture, the same MoonViT 400M vision encoder, the same mandatory thinking mode, and produces identical quality outputs. The only difference is how fast those outputs arrive.

Released on June 15, 2026 - three days after the base K2.7 Code release on June 12 - HighSpeed was Moonshot's direct response to developer feedback that agentic coding workflows are bottlenecked by model throughput. When an AI agent runs hundreds of steps autonomously, each token generated in chain-of-thought reasoning and code output adds latency that compounds across the entire session. At 180–260 tokens per second, those bottlenecks largely disappear.

Access is rolling out to Kimi Code Beta Program members, Kimi API developers, and Kimi Business users in order. No invite is required - joining the Beta Program gives you access as capacity becomes available. Moonshot has stated they are actively expanding infrastructure to serve more users.

  • Identical model to K2.7 Code - same weights, same quality, same capabilities
  • ~180 tok/s on median-length coding inputs
  • Up to 260 tok/s on shorter-context tasks
  • 6× faster than the standard K2.7 Code serving tier
  • API model ID: kimi-k2.7-code-highspeed
  • Currently limited capacity - expanding as resources increase
⚡ JUNE 15, 2026
Instant.
Agentic.
Open.
High-throughput variant of K2.7 Code. Same model, extreme speed. Designed for agentic workflows where token throughput determines task completion time.
260Peak tok/s
Speed boost
1TParameters
02 - K2.7 CODE ARCHITECTURE

The Model Behind HighSpeed

Kimi K2.7 Code - the model HighSpeed serves at extreme throughput - was released on June 12, 2026 as Moonshot AI's most capable coding model to date. It is a coding-focused, agentic successor to K2.6, built on the same 1T-parameter MoE architecture but retrained with a new reward model and data pipeline optimized specifically for long-horizon software engineering tasks.

The architecture retains what made K2.6 exceptional: 1 trillion total parameters with 32B active per token across 384 experts (8 selected + 1 shared) in 61 layers. Multi-Head Latent Attention (MLA) handles long contexts efficiently; SwiGLU activations power the feed-forward path. The MoonViT 400M vision encoder adds full multimodal capability - text, image, and video input are all supported natively.

What changed from K2.6: K2.7 Code was specifically tuned to reduce overthinking by ~30% - its reward model was trained to produce correct code with fewer intermediate reasoning tokens, without sacrificing output quality. This makes it simultaneously more efficient and faster, even before the HighSpeed serving infrastructure is applied.

  • 1T total / 32B active - MoE sparse activation per token
  • 384 experts, 8 + 1 shared per token - fine-grained expert routing
  • 61 layers - 1 dense + 60 MoE layers
  • MLA attention + SwiGLU - efficient long-context processing
  • MoonViT 400M encoder - native vision for images and video
  • 256K context window - consistent with K2.6
Release note

K2.7 Code is currently available on the official Kimi API and on Cloudflare Workers AI (@cf/moonshotai/kimi-k2.7-code). Model weights are on HuggingFace under a Modified MIT License. Existing K2.6 deployment patterns (vLLM, SGLang, KTransformers) work with K2.7 Code with a simple model ID swap.

⚙ K2.7 CODE · ARCHITECTURE
1T Params.
32B Active.
Vision.
Same MoE foundation as K2.6. Retuned reward model for coding. MoonViT 400M for multimodal. 30% fewer reasoning tokens by design.
384Experts
256KContext
400MVision encoder
03 - MANDATORY THINKING MODE

Always-On Reasoning - By Design

Kimi K2.7 Code - and by extension K2.7 Code HighSpeed - operates with thinking mode permanently enabled. Unlike K2.6, which supported a non-thinking instant mode, K2.7 Code removes this option entirely. The API will return an error if you attempt to disable thinking. This is a deliberate design choice, not a limitation.

Moonshot's rationale: for the long-horizon software engineering tasks K2.7 Code targets, the reasoning chain is load-bearing. Allowing it to be skipped for "fast" responses introduced instability in agentic multi-step workflows - the model would take shortcuts that became evident only after several hundred steps. The mandatory thinking guarantee means every K2.7 Code response is backed by the same quality of reasoning, regardless of task length.

The tradeoff is straightforward: you cannot make cheap, no-reasoning calls to K2.7 Code. Every request reasons first. However, the 30% reduction in reasoning tokens versus K2.6 significantly narrows this cost premium - and HighSpeed's throughput advantage ensures those reasoning tokens arrive faster than ever.

  • Thinking mode always on - cannot be disabled via API parameter
  • Temperature locked at 1.0, top_p 0.95 - server-side, not adjustable
  • Preserve-thinking across turns - reasoning chain retained between conversation turns
  • Default output cap: 32,768 tokens - configurable up to context limit
  • Access reasoning via reasoning_content - full chain-of-thought in every response
  • 30% fewer reasoning tokens vs K2.6 - efficiency built into the model, not the mode
Migration note from K2.6

If your K2.6 integration used non-thinking (instant) mode with thinking: False in extra_body, this parameter is ignored on K2.7 Code - the model always thinks. Review your token budget assumptions before migrating production workloads, as every request will incur reasoning token usage. Use the reported 30% efficiency gain to estimate your revised cost model.

🧠 THINKING MODE · ALWAYS ON
Reason
Every
Time.
Thinking is mandatory - by design. Every request reasons first. Preserve-thinking keeps the chain across multi-turn sessions. 30% fewer tokens than K2.6 means the premium is lower than ever.
−30%vs K2.6 thinking
1.0Temperature
32KDefault output
04 - KEY FEATURES

What Makes K2.7 Code HighSpeed Different

HighSpeed is more than a faster server - it combines the throughput breakthrough with K2.7 Code's fundamental improvements over K2.6 in coding quality, tool use, and token efficiency.

// 01 · THROUGHPUT

Up to 6× Token Speed

~180 tok/s on median coding tasks, up to 260 tok/s on shorter-context inputs. Delivered by Moonshot's optimized HighSpeed serving infrastructure - not a different model, not a lighter variant. Same quality, dramatically faster delivery.

180–260 tok/s · Production grade
// 02 · EFFICIENCY

−30% Reasoning Token Usage

K2.7 Code produces correct code with ~30% fewer thinking tokens than K2.6 on average. For agentic sessions running hundreds of steps, this compounding saves translate to meaningfully lower API costs and faster completion times - before HighSpeed throughput is even applied.

30% efficiency gain · Compounds at scale
// 03 · MCP TOOL USE

SOTA on MCP Mark Verified

K2.7 Code scores 81.1 on MCP Mark Verified - beating Claude Opus 4.8's 76.4. Model Context Protocol tool use is a primary focus: correct invocation of Notion, GitHub, Filesystem, Postgres, and Playwright tools in CI/CD and agent loop workflows.

81.1 MCP Mark · Beats Opus 4.8
// 04 · MULTIMODAL

Text, Image, and Video Input

MoonViT 400M encoder enables native multimodal input. Upload screenshots, design mockups, diagrams, or video recordings alongside code and documentation. K2.7 Code reads all of them simultaneously in a single prompt - critical for UI debugging and visual debugging workflows.

MoonViT 400M · Text + Image + Video
// 05 · CONTEXT

256K Token Context Window

256K tokens (~192,000 words) of context - equivalent to a large enterprise codebase or an entire library of documentation. Shared across all K2.7 Code access paths. HighSpeed preserves the full context length; no degradation from the standard model.

256K context · Unchanged from K2.7
// 06 · OPENAI COMPATIBLE

Drop-In API Compatibility

Kimi's API is fully OpenAI-compatible. Change two lines in any existing integration: base_url and model. The HighSpeed model ID is kimi-k2.7-code-highspeed. Streaming, function calling, tool use, and structured outputs all work identically to OpenAI's format.

OpenAI-compatible · Two-line migration
05 - BENCHMARKS

K2.7 Code Performance vs K2.6

Moonshot published six benchmark results comparing K2.7 Code against K2.6, GPT-5.5, and Claude Opus 4.8. Important context: every benchmark listed here is a Moonshot proprietary suite - Kimi Code Bench v2, Program Bench, MLS Bench Lite, MCP Atlas, MCP Mark Verified, and Kimi Claw 24/7 Bench are all designed and administered by Moonshot. No independent results on SWE-bench Verified, LiveCodeBench, GPQA Diamond, or Terminal-Bench 2.0 had been published at launch. Use these numbers as directional indicators, not independently verified scores.

Benchmark K2.7 Code vs K2.6 GPT-5.5* Claude Opus 4.8* Visual
Kimi Code Bench v2 62.0 +21.8% 69.0 67.4
Program Bench 53.6 +11.0% - -
MLS Bench Lite 35.1 +31.5% 35.5 42.8
MCP Atlas 76.0 ~+10% - -
MCP Mark Verified 81.1 ~+10% - 76.4 (K2.7 leads)
Kimi Claw 24/7 Bench ~+10% vs K2.6 +10% - -

*GPT-5.5 ran in Codex xhigh mode; Claude Opus 4.8 ran in Claude Code xhigh mode; K2.7 Code ran in Kimi Code CLI. Different compute configurations - comparisons are directional only. All benchmarks are Moonshot proprietary. As of June 15, 2026, no independent SWE-bench Verified, LiveCodeBench, or GPQA Diamond results exist for K2.7 Code.

Benchmark transparency

Every published benchmark for K2.7 Code is run by Moonshot on Moonshot-designed suites. This is common practice but worth acknowledging explicitly. Test K2.7 Code against your own actual workloads before drawing conclusions from the numbers above. The 30% token efficiency gain is the metric most likely to reflect consistently across diverse real-world tasks.

06 - API & INTEGRATION

Using Kimi K2.7 Code HighSpeed

The HighSpeed variant uses an identical API surface to K2.7 Code standard - the only change is the model string. Any existing K2.7 Code integration can switch to HighSpeed by changing one line. The API lives at platform.kimi.ai and is fully OpenAI SDK compatible.

Python · OpenAI SDK · K2.7 Code HighSpeed
# pip install openai from openai import OpenAI client = OpenAI( api_key="YOUR_KIMI_API_KEY", # platform.kimi.ai base_url="https://api.moonshot.ai/v1" ) # HighSpeed - same quality, 6x faster delivery response = client.chat.completions.create( model="kimi-k2.7-code-highspeed", # ← the only change vs standard temperature=1.0, # locked server-side anyway max_tokens=32768, messages=[ {"role": "system", "content": "You are a senior software engineer."}, {"role": "user", "content": "Refactor this authentication module for async/await..."} ] ) # Access thinking chain (always present in K2.7 Code) print("Reasoning:", response.choices[0].message.reasoning_content) print("Output:", response.choices[0].message.content) # Preserve thinking across turns (multi-turn agentic loop) messages = [] for step in agent_steps: messages.append({"role": "user", "content": step}) resp = client.chat.completions.create( model="kimi-k2.7-code-highspeed", messages=messages, max_tokens=32768 ) # Add assistant message WITH reasoning_content to preserve thinking msg = resp.choices[0].message messages.append({ "role": "assistant", "content": msg.content, "reasoning_content": msg.reasoning_content # ← preserve this })

Access Paths

// KIMI API

Official API - platform.kimi.ai

Direct access to HighSpeed via Moonshot's API. Get your key at platform.kimi.ai. Model ID: kimi-k2.7-code-highspeed. Standard K2.7 Code: kimi-k2.7-code. Cached input: $0.19/1M. Miss input: $0.95/1M (standard).

Primary access path
// KIMI CODE CLI

Kimi Code Beta Program

HighSpeed is rolling out to Kimi Code Beta Program members first. Join at kimi.com/code/beta - no invite required. Members in the Beta Program get access as capacity becomes available. The Kimi Code CLI uses K2.7 Code as its default model as of June 12, 2026.

Beta Program · No invite needed
// OPENROUTER

OpenRouter - Multi-Provider

K2.7 Code is available on OpenRouter at ~$0.72/1M input · $3.49/1M output via 13+ providers. Routes automatically for best uptime. HighSpeed variant availability on OpenRouter depends on individual provider support.

13+ providers · Auto-routing
// CLOUDFLARE WORKERS AI

Workers AI - Edge Deployment

K2.7 Code is available on Cloudflare Workers AI as @cf/moonshotai/kimi-k2.7-code. Use via Workers AI binding, REST API, or OpenAI-compatible endpoint. Cached input: $0.19/1M.

Edge · Workers AI binding
07 - PRICING

K2.7 Code API Pricing

Kimi K2.7 Code HighSpeed is priced at a premium over the standard K2.7 Code tier - reflecting the dedicated high-throughput serving infrastructure. Both variants significantly undercut closed-source models at equivalent capability levels. Kimi Code membership plans start at $19/month and provide included model access without per-token billing.

K2.7 Code Standard
kimi-k2.7-code
$0.95per 1M input tokens
$4.00per 1M output tokens
$0.19per 1M cached input
  • Includes
  • Full K2.7 Code capability
  • 256K context window
  • MoonViT multimodal input
  • Thinking mode (mandatory)
  • Prompt caching support
  • OpenAI + Anthropic compatible
  • Standard throughput
⚡ K2.7 Code HighSpeed
kimi-k2.7-code-highspeed
$1.90per 1M input tokens
$8.00per 1M output tokens
$0.19per 1M cached input
  • Everything Standard +
  • ~180 tok/s median coding
  • Up to 260 tok/s peak
  • 6× faster than standard tier
  • Real-time agentic streaming
  • Priority compute allocation
  • Limited capacity - expanding
Cost model for agentic workflows

For long agentic sessions where reasoning tokens dominate cost, K2.7 Code's 30% reduction in thinking tokens offsets a significant portion of the HighSpeed premium. A session that previously consumed ~2 million reasoning tokens with K2.6 will consume roughly ~1.4 million with K2.7 Code. At HighSpeed pricing ($8.00/1M output), that's a saving of ~$4.80 per session - before accounting for the time value of faster completion. Run the math against your actual workload to determine which tier optimizes your specific cost-speed trade-off.

08 - USE CASES

Where K2.7 Code HighSpeed Shines

The HighSpeed variant is most valuable when token throughput is a meaningful bottleneck - typically agentic workflows with many steps, interactive developer tools, and any context where wait time directly impacts workflow efficiency.

🤖 Autonomous Coding Agents

Multi-step agentic sessions (debugging, refactoring, feature implementation) run faster start-to-finish. At 180 tok/s, a 1,000-token reasoning chain arrives in ~5.5 seconds vs ~30s at standard speed.

🔁 CI/CD Pipeline Integration

Automated code review, test generation, and PR analysis in CI pipelines. HighSpeed ensures K2.7 Code completes before pipeline timeouts - critical for synchronous review gates.

💬 Interactive Coding Sessions

Real-time pair programming in Kimi Code CLI or IDE integrations. At sub-second first-token latency, the interaction feels like a live pair rather than a waiting room.

🔧 MCP Tool Orchestration

Workflows invoking MCP tools (GitHub, Notion, Filesystem, Postgres) with many sequential calls benefit directly - each step's reasoning and output arrives faster, reducing total session time.

📊 Batch Code Analysis

Processing large codebases file-by-file or function-by-function in parallel agentic workflows. HighSpeed makes per-file analysis fast enough to complete full repository scans in minutes.

🌐 Real-Time Code Review

Synchronous review of commits and PRs as they land. HighSpeed throughput enables < 10 second analysis turnaround on typical commit diffs - fast enough for blocking developer workflows.

09 - K2 MODEL FAMILY TIMELINE

One Year of Kimi K2 Evolution

From the original K2 release in July 2025 to K2.7 Code HighSpeed in June 2026 - Moonshot AI shipped six major updates in twelve months, consistently iterating faster than any closed-source frontier lab over the same period.

K2
July 2025

Kimi K2 - Open-Source Frontier Foundation

1T MoE, 32B active, 128K context, MuonClip optimizer, zero training instability on 15.5T tokens. Modified MIT License. Sets the open-source agentic baseline. SWE-bench: 65.8%.

K2-T
November 2025

K2 Thinking - Interleaved Reasoning + Tool Use

Interleaved chain-of-thought and native tool calls, up to 300 sequential steps. Native INT4 QAT for 2× speed. Tencent CodeBuddy integrates as core model. temperature=1.0.

K2.5
January 27, 2026

Kimi K2.5 - Visual Agentic Intelligence

MoonViT 400M for native multimodal (images + video). 256K context. Agent Swarm v1: 100 parallel sub-agents, 4.5× speedup. SWE-bench 76.8%. Cursor Composer 2 built on K2.5.

K2.6
April 20, 2026

Kimi K2.6 - Long-Horizon Agentic Coding GA

262K context. Agent Swarm v2: 300 sub-agents, 4,000 steps. Claw Groups. Document-to-Skill. SWE-bench 80.2%, BrowseComp Swarm 86.3%. Non-thinking (instant) mode available.

K2.7
June 12, 2026

Kimi K2.7 Code - Coding-Specialist Refresh

Coding-focused retraining of K2.6. Mandatory thinking mode. −30% reasoning tokens vs K2.6. +21.8% Kimi Code Bench v2. MCP Mark 81.1 (beats Opus 4.8). Open weights HuggingFace.

HS
June 15, 2026 - Now Available

Kimi K2.7 Code HighSpeed - 260 tok/s · 6× Faster

Same K2.7 Code model, optimized high-throughput serving infrastructure. ~180 tok/s median, up to 260 tok/s peak. Rolling out to Kimi Code Beta, API developers, Kimi Business. API: kimi-k2.7-code-highspeed.

10 - COMPARE

K2.7 Code HighSpeed vs the Field

Positioned as the open-weight speed leader for coding: frontier-competitive quality at a fraction of closed-source cost, with the highest raw throughput of any publicly available coding model as of June 2026.

Model
K2.7 Code HighSpeed
Moonshot AI · Jun 2026
K2.7 Code Std
Moonshot AI
GPT-5.5
OpenAI
Claude Opus 4.8
Anthropic
K2.6
Moonshot AI
Speed & Throughput
Token throughput (approx)180–260 tok/s~30–50 tok/s~60–80 tok/s~50–70 tok/s~40–60 tok/s
Speed vs K2.7 standard6× faster~1.5×~1.3×
Model Capabilities
Parameters (total / active)1T / 32B1T / 32B~200B~200B1T / 32B
Context window256K256K400K (1M Pro)200K262K
Vision / multimodal✓ MoonViT 400M
Thinking modeAlways onAlways onOptionalOptionalSwitchable
Open weights✓ Modified MIT
Pricing (approximate)
Input $/1M tokens$1.90$0.95$5.00$15.00$0.60
Output $/1M tokens$8.00$4.00$30.00$75.00$4.00
vs GPT-5.5 output cost3.75× cheaper7.5× cheaper-2.5× more7.5× cheaper
Key Benchmarks (Kimi Code Bench v2 - vendor-reported)
Kimi Code Bench v262.062.069.0*67.4*50.9
MCP Mark Verified81.181.1-76.4 (lower)~73

*Competitor benchmark scores from Moonshot's first-party table - GPT-5.5 in Codex xhigh mode, Opus 4.8 in Claude Code xhigh. Not directly comparable to standard API mode. All Kimi Code Bench results are Moonshot proprietary. Throughput figures are approximate and vary by task length and load.

11 - LIMITATIONS & HONEST CAVEATS

What to Know Before Deploying

// 01

All Benchmarks Are Proprietary

Every published benchmark for K2.7 Code - Kimi Code Bench v2, Program Bench, MLS Bench Lite, MCP Atlas, MCP Mark Verified - is a Moonshot-designed and Moonshot-administered suite. No SWE-bench Verified, LiveCodeBench, Terminal-Bench 2.0, or GPQA Diamond results exist as of June 15, 2026. Treat published scores as vendor-reported and directional.

First-party only
// 02

Capacity Limited During Rollout

Moonshot explicitly noted that HighSpeed capacity is constrained at launch. The experience may fluctuate - throughput could vary significantly under high load, and access may be rate-limited more aggressively than standard. Capacity is actively expanding, but plan for variability in production until the rollout stabilizes.

Limited capacity initially
// 03

No Non-Thinking Mode

Thinking mode is permanently enabled. Every request reasons before responding - there is no cheaper, faster path for simple or trivial queries. For teams that used K2.6's instant mode for lightweight calls, this represents a meaningful cost increase. The 30% efficiency gain helps but doesn't fully offset for simple workloads.

Always reasons first
// 04

256K Context vs K2.6's 262K

K2.7 Code has a 256K context window - slightly shorter than K2.6's 262K. For the overwhelming majority of workflows this is irrelevant, but teams whose sessions reliably approach 260K+ tokens should account for this regression or maintain a K2.6 fallback for extremely long-context tasks.

256K vs 262K in K2.6
12 - FAQ

Frequently Asked Questions

Build at the Speed of Thought

Kimi K2.7 Code HighSpeed delivers up to 260 tokens per second - the fastest serving tier for any open-weight frontier coding model. Join the Beta Program or access via API today.